Case selection and the justification behind it are crucial elements of research. In my case, my case selection was vulnerable because I have chosen my own country. Although it is not forbidden, my task was to provide a purely scientific justification why Ukraine and not any other constitutes an interesting case to study. In other words, the fact that a scholar has chosen his or her own country is not a sufficient justification.
Analysis of the previous studies of any topic is a crucial element of a research process. As a rule, it should be done even before the actual choice of the topic. In order to make a choice about the topic among several alternatives, a researcher should do at least preliminary research about the previous studies within a shortlist of topics to choose. While doing this, a researcher should ask the questions how the variables studied were conceptualized and operationalized, what were the results of these studies, are they valid and reliable, what is the unresolved research dilemma he/she will be dealing with, how can he/she contribute to the studies on this topic, does he/she have enough data and methodological tools to complete the research.
For example, Enyedi in his study found that party system institutionalization was a cause of de-democratization of Hungary by arguing that populist polarization has emerged partly because of the high level of party system institutionalization (Enyedi, 2016, p. 218). The case of Hungary can be considered as a deviant one because the scholar agrees that in general party system institutionalization is conducive to democratization.
As one of the examples of the regional studies could serve the study of Latin American countries by Dix. Regarding the findings of the study, it was argued that on average there was a modest institutionalization of the party system in Latin America between the 1960s and 1980s which had the potential to positively influence the level of democracy in this region (Dix, 1998, p. 488). As for the critique of the study, it should be mentioned that the author conflated the concepts of the institutionalization of the party system with the institutionalization of parties without explaining why these concepts should be used interchangeably. Moreover, regarding some indicators the author has not had empirical data which resulted in judgemental arguments regarding them.
As for the measurement of democracy, the authors use the Polity II democracy index. Also, several control variables were used. The study has found that high level of electoral volatility is not associated with a high level of democracy which supports the theoretical argument about the association between party system institutionalization and the level of democracy (Thames and Robbins, 2006, p. 10).
Party system institutionalization is one of the two core variables in this study along with the level of democracy. However, contrary to the general principles of the research chain, in this case it is almost impossible to conclude which of the variables is independent one and which is dependent. For example, it can be argued that party system institutionalization caused changes in the level of democracy. However, it can also be stated that certain level of democracy led to certain level of party system institutionalization. Therefore, the question of causality is beyond the scope of this particular study. The goal is merely to prove an association between phenomena which means proving there is a stable pattern of mutual dynamics of these variable between 1994 and 2007 in Ukraine. Last, but not the least, causality might not even be attainable within this research question because of complex interrelations and mutual influences of phenomena in social and political realms.
The concept of volatility is present in this study in order to double-check the two operationalization of party system institutionalization and find out if they lead to the same conclusions regarding the association between party system institutionalization and democratization in Ukraine. Another reason is that the electoral volatility is still a dominant operationalization of party system institutionalization whereas party system closure is a relatively new measurement which has not gained that much popularity despite its arguable advantages.
Electoral democracy consists of such criteria as responsiveness to the citizens, electoral competition without "fraud and systemic irregularities", free operation of political and civil society organizations, the impact of elections on the composition of the main executive institution in a political system, freedom of expression and independence of media between elections (University of Gothenburg, V-Dem Institute, 2018, p.40). Moreover, the index of electoral democracy is a basis for all other indices by V-Dem and is taken into account in the calculations of all the varieties of democracy.
Liberal democracy is characterized by the protection of individual and minority rights, limits placed on government such as checks and balances, the independent judiciary and the rule of law (University of Gothenburg, V-Dem Institute, 2018 pp.40-41).
Deliberative democracy emphasizes the importance of a rational dialogue based on the principle of the common good as opposed to emotional arguments. The deliberative principle of democracy essentially means that.
Participatory democracy takes electoral participation for granted and is thereby focused on the participation of citizens in civil society organizations, elements of direct democracy and the powers of existing subnational governmental institutions (University of Gothenburg, V-Dem Institute, 2018 pp. 40-41).
Egalitarian democracy consists of three essential criteria: "1 rights and freedoms of individuals are protected equally across all social groups; and 2 resources are distributed equally across all social groups; 3 groups and individuals enjoy equal access to power" (University of Gothenburg, V-Dem Institute, 2018 pp.42

PARTY SYSTEM INSTITUTIONALIZATION AND DEMOCRATIZATION OF A POLITICAL REGIME IN UKRAINE. PAPER WITH METHODOLOGICAL EXPLANATIONS

Oleksandr Kuchynskyi
Student, Johan Skytte Institute of Political Studies, University of Tartu
instructor: Martin Mölder
researcher
RESEARCH PROBLEM

The paper is going to answer the question if there is an association between the dynamics of party system institutionalization in Ukraine and the level of democracy in Ukraine. In other words, it will be examined whether the processes of institutionalization and deinstitutionalization have an association with the processes of democratization and de-democratization, respectively.

Party system institutionalization is generally regarded as conducive to democratization. Conversely, noninstitutionalized party systems generally do not prove themselves able to reach a considerable level and quality of democracy (Thames and Robbins, 2007, p.3). In particular, the high frequency of changes within the party system severely hinders the accountability of political parties (Thames and Robbins, 2007, p.3). Moreover, the lack of predictability within the party system may cause frustration among citizens and make them unwilling to participate in politics. However, this tendency works much better in the extreme rather than in-between cases which are neither extremely institutionalized nor extremely noninstitutionalized (Thames and Robbins, 2007, p.3).

The case of Ukraine was chosen because this country has one of the highest levels of volatility and fragmentation within Europe (Casal Bertoa 2016), but the association between the changes in the level of party system institutionalization and the democratization of the country has not yet been examined *. Moreover, two major shifts that happened in the party system took place after the revolutions of 2004-2005 and 2013-2014 and were regarded as manifestations of democratization although these revolutions also led to the increase of the volatility and which contradicts the prevailing hypothesis among the scholars about the relationship between the party system institutionalization and democratization (McFaul, 2007, pp. 80-83; Fedorenko, Rybiy and Umland, 2016, p. 609). I am going to test whether these two cases related to the revolutions were exceptional in their contradiction to the theoretical hypotheses of the majority of scholars and what is the association between the level of democracy and the level of party system institutionalization in Ukraine from 1994 to 2017.
PREVIOUS STUDIES ON THE TOPIC

There have been many studies on the topic of the link between party system institutionalization and democratization*. Mostly these were single case-studies (Enyedi, 2016)* and comparative small-n regional studies (Dix, 1998)*.

As for the large-n studies, the study by Thames and Robbins carried out a study using a cross-sectional dataset of 58 countries from 1974 to 2005. The authors operationalized party system institutionalization as electoral volatility lagged by one year (Thames and Robbins, 2006, p. 10)*.
CENTRAL CONCEPTS AND THEIR MEASUREMENT

Party system institutionalization *

There is a relative consensus regarding the basic feature of party system institutionalization — stabilization. For example, one general definition of this kind is provided by Huntington: "process by which organizations and procedures acquire value and stability" (1968, p.12). However, there are different points of view regarding what this stability is. In particular, one group of scholars is concentrated on the stability of the political parties themselves (Hicken and Martínez Kuhonta, 2011, p. 573). Others are more interested in the dynamics of interactions of the parties within the system. For instance, Casal Bertoa defined party system institutionalization as "...the process by which the patterns of interaction among political parties become routine, predictable, and stable over time"(2014, p. 17). In my opinion, the emphasis on interactions, but not on the parties as such is an advantage of the definition by Casal Bertoa. Although parties' institutionalization is also important for the institutionalization of the system, the level of institutionalization of the parties can be assumed from the institutionalization of the interactions because stable interactions are possible only between stable parties. Another similar, but a more parsimonious definition is "there is stability in who the main parties are and how they behave" (Mainwaring 1999, p. 25).

Two different logics of conceptualizing party system institutionalization had led to two different logics of its operationalization. For example, electoral volatility is concentrated on the extent to which the electoral results of parties are stable from election to election. As for the justification of the association between the volatility and the level of democracy, it is argued that high levels of volatility are not conducive to democracy since the electorate is agnostic about the parameters of competition and cannot have any predictions about the results since the victory of political outsiders is likely in that system (Mainwaring and Zoco, 2007, pp.157-158). Such an approach is a manifestation of the focus on the institutionalization of the parties themselves which could be assumed from the changes in their electoral results. The adherents of a different logic do not abandon the latter one at all, but rather enrich it. Namely, the creators of the index of party system closure derive the information about it from the changes in the shares of ministries between the parties forming a coalition in a particular country (Casal Bertoa and Enyedi, 2016, p. 5-6). The main focus is on the stability of coalition formation patterns which are reflected in the share of ministries by party affiliations. However, the attention to the relative strength of the parties is not abandoned since the share of votes that the party received during the last parliamentary elections is usually the strongest argument during the negotiation processes regarding the number of ministries a particular party would receive.
INDICATORS OF PARTY SYSTEM INSTITUTIONALIZATION

Volatility *

Electoral volatility measures the net change in the voting patterns from election to election (Pedersen, 1979, p.3). Volatility also takes into account the splits and mergers among the parties. According to Pedersen, volatility is measured by the following formula:

"TEV=Σ|Vi,t-1-Vi,t|/2"

in which Vi,t is the vote share for a party at a given election (t) and Vi,t-1 is the vote share of the same party ith at the previous elections (t-1)" (Pedersen, 1979, p.4).

Party System Closure

The concept of party system closure is concentrated on the most consequential element of the party competition — the composition of government. Thus, it uses one type of information for the three components (access, alternation, and formula) — the shares of ministries in a government by party affiliation (Casal Bertoa and Enyedi, 2016, p. 5-6). The logic behind the emergence of this concept and the corresponding index of party system closure is that the shares of ministries in the government among the parties does not equal the share of the votes or seats that the parties obtained as a result of the elections because the factor of the interactions between the parties in the forms of negotiations and bargaining comes into play.

Access

Bertoa and Enyedi operationalize access to government as the percentage of parties in the government which have ever governed before (Casal Bertoa and Enyedi, 2016, p.6).

Alternation

The calculation of the alternation in government is based on the logic of Pedersen's index of ministerial volatility with the difference that the percentage of votes for a particular party is replaced with the percentage of ministries governed by the representatives of political parties for two consecutive elections (Whogoverns.eu, 2015, pp. 1-2). To calculate alternation, all the differences in the percentage of ministers belonging to governing parties should be added and divided by two. It should also be noted that since both variations of the extreme scores (close to 0 and close to 100) are manifestations of wholesale alternation, the score for alternation should be standardized according to the following formula:

IGA = (MVis-50)*2

where IGA is alternation and MVis is ministerial volatility (Whogoverns.eu, 2015, p. 2). However, the score should be standardized if and only if the initial score for alternation is less than 50.

The authors consider both total alternation and no alternation as the indicators of party system closure. In the case of no alternation, the reasoning is intuitively understandable because the composition of government does not change at all. In the case of total alternation, the argument is that in this case there is no continuity in the party composition between the governments (Casal Bertoa and Enyedi, 2016, p.4).

Formula

The calculation of formula is more complicated than the calculation of two other indicators since it differs according to the type of the previous government: single-party or multi-party. (Bertoa and Enyedi, 2014, pp. 5-6). In case only one or few parties from the previous government coalition are present in the current government, this coalition should be compared to the most similar composition of parties in the coalition which should not necessarily be the last government, but the most similar to the current one. In this case, the innovativeness of the formula is calculated by the percentage of the ministries in the government being analyzed belonging to the parties present in both coalitional governments being compared. If there are several governments having the same number of parties present in the current government, then the comparison should be made with the previous government closest in time. If the government was previously a single-party one, the formula would be the percentage of the ministries belonging to the party which previously governed alone after the emergence of new parties in the coalition. On the contrary, if the government composition changed from being multiparty to a single-party, then the governing formula would be the percentage of the ministers that the party which is currently governing alone had in the previous coalitional government (Whogoverns.eu, 2015, p. 2).

Such an operationalization allows calculating the composite score of closure which would be the average percentage of all three indicators (Casal Bertoa and Enyedi, 2016, p. 6). It should also be noted that if several changes in government happened during a calendar year, the average score for all the changes should be calculated (Whogoverns.eu, 2015, p. 3). Similarly, in order to calculate the value of closure for all the years of a democratic regime, the average value for all the years should be calculated.

Both extreme closure and openness have disadvantages in terms of democracy and the efficiency of government, but for different reasons. Having extreme openness means that the interactions between the parties are not predictable which means total instability both of the party systems and the governments. On the other hand, extreme closure means that a country is basically governed by a single party or a coalition of parties that does not change its composition. However, as Bertoa and Enyedi argue, closed systems tend to facilitate democracy more (2014, p. 11). They also conducted statistical analysis and found that the differences between the cases of survival and collapse between the open and closed systems were statistically significant. Moreover, even within the democratic countries, extreme openness at the beginning of the regime existence is conducive to the breakdown of the regimes. In particular, the study was carried out by Berg-Schlosser and Mitchell in which they explained the survival or collapse of democracy by the fact whether the party system was open or closed in the inter-war period (Bértoa and Enyedi, 2016, p. 8).

Democracy

Since the aim of this paper is to find an association between the changes in the level of party system institutionalization and the level of democracy, then the concept of democracy itself needs to be conceptualized and operationalized. In this study, am going to use V-Dem database as an indicator of the changes in the level of democracy. The reason is that this index does not offer any single definition of democracy, but calculates the scores of democracy according to different theoretical frameworks of democracy: electoral democracy*, liberal democracy*, deliberative democracy*, participatory democrac*, egalitarian democracy*.
OVERVIEW OF THE SOURCES OF DATA

  • Shares of ministries between political parties is obtained from the database WhoGoverns until 2013, but the newer information regarding the composition of governments and party affiliations of ministers can be found in Ukrainian media articles (Pravda.com.ua, 2014; dw.com, 2014; www.depo.ua, 2016).

  • Scores for party system closure in Ukraine was calculated by myself due to the lack of publicly accessible calculations done by other scholars.

  • The information about electoral volatility from 2002 to 2012 is taken from the study conducted by Rybiy(2013, p. 411), and for 2014 from the study conducted by Razumkov center (2017, p.61). It should be also noted that for the elections in 1994 and 1998, the electoral volatility was impossible to calculate since Ukraine had a majoritarian electoral system, and the party affiliation of the majority of the MP was impossible to identify regarding the majority of the MPs(Rybiy, 2013, p.408).

  • As for the operationalization of democracy, V-Dem database is used.

ANALYSIS AND RESULTS

In general, the analysis of the relationship between the indicators of party system institutionalization (electoral volatility and party system closure) conforms with the theoretical arguments regarding the association between party system institutionalization and democratization.

Electoral volatility

As for the volatility, in most cases, there is a negative association with the change in the level of democracy. Namely, between 2006 and 2007 the level of democracy had increased which was accompanied by the decrease in the level of volatility which corresponds to the theoretical framework discussed earlier. However, the alternative explanation might be that volatility decreased since the snap elections which took place in 2007 were held only year after the elections in 2006 so that the preferences of the voters and the party system at large might not change as significantly as it happened between 2002 and 2006. In addition, it should be taken into account that the Orange Revolution happened in 2004 could had contributed to the change in the balance of power between the parties. After that, between 2007 and 2012, and between 2012 and 2014, the increase in volatility was accompanied by the decrease in the level of democracy according to V-Dem. However, between 2002 and 2006, the increase in the level of volatility happened simultaneously with the increase in the level of democracy which contradicts the theoretical explanations discussed earlier. Probably, such an inconsistency may again be explained with the factor of the Orange Revolution in 2004 which contributed to the changes in the party system. In general, the exceptional cases of positive association between electoral volatility and the level of democracy do not refute the argument that a high level of electoral volatility is not conducive to the increase in the level of democracy and vice versa. Moreover, the impact of revolutions leading to deep transformations of political regimes and the fact that they are rather exceptional rather than a routine phenomenon in a political process encourages to treat the influence of this factor seriously and not conclude that the whole explanatory logic does not work because of one case that does not conform to it.


Graph 1. The association between the electoral volatility and varieties of democracy according to V-Dem between 2002 and 2014. (For the illustrative purposes, the values for the varieties of democracy were transformed to the same scale by multiplying them by 10. The same applies to the following graphs) // Oleksandr Kuchynskyi ©️

Party system closure

The interpretation of the results for the party system closure greatly depends on the time perspective taken by me. For instance, if a year-by-year perspective is taken, there is hardly an association between the dynamics of party system closure and the dynamics of democracy in Ukraine. Let us have a brief look at the yearly dynamics of party system closure in Ukraine and the dynamics of democracy.

From 1994 to 1996 it is evident that the score for party system closure dropped from 61,1 to 40,4, but in both cases the level of democracy rose. From 1996 to 1998 the index of closure had risen from 40,4 to 72,1 along with the decrease in the level of democracy. From 1999 to 2000 the level of closure decreased whereas the level of democracy stayed stable. Between 2000 and 2002, the index of closure fell dramatically (from 100 to 42,7), but the level of democracy stayed almost the same. Such yearly examination could be continued to 2017, but, as it is evident from the graphs, the yearly dynamics of party system closure does not explain the yearly dynamics of democracy in Ukraine.

However, in the case of party system closure it is much more useful not to focus on the separate years, but to have a look on more long-term tendencies. The reason is that the yearly changes in government do not provide an information regarding the dynamics of democracy in a country. For instance, a particular year may be unstable, with several major changes in the composition of government, but at the same time many democratic transformations may happen during that year. On the contrary, there can be no changes in the composition of government within a year or even several years, but it does not necessarily mean the progress of democracy. For example, Ukraine had maximum values of closure during the presidency of Yanukovych, the period which was rather an instance of democracy deterioration. Therefore, the time perspective for the assessment of the explanatory power of party system closure to determine the dynamics of democracy should be considerably longer. Moreover, coming back to the theoretical discussions by Casal Bertoa who created the index of party system closure as an indicator of party system institutionalization, party system is a process of stabilization of patterns of interactions between parties that happens over a certain period of time and it can be assumed that this period of time is considerably longer than a year (Casal Bertoa, 2014, p. 17). However, the question remains, how long the time perspective of the analysis should be in order to explain the dynamics of the level of democracy.


Graph 2. Association between the index of party system closure and varieties of democracy according to V-Dem between 1994 and 2017 // Oleksandr Kuchynskyi ©️
For instance, on the Graph 3, the annual values for party system closure are grouped into 5 periods, 4 years each in order to analyze if more general changes in party system closure correspond to the changes in the level of democracy. However, the changes in the the level of closure do not explain the changes in the level of democracy. The level of closure between 1998-2001 increased in comparison with 1994-1997, but during the same period the level of democracy fell down. However, between 2002-2005, the increase in the level of closure corresponded to the increase in the level of closure which is consistent with the theoretical explanations. Between 2010 and 2013, the level of closure increased, but the level of democracy fell down. Therefore, the time perspective is not long enough.
Graph 3. The dynamics of party system closure by 4-year periods // Oleksandr Kuchynskyi ©️
Taking a much longer perspective of analysis finally allows to evidence the association between the variables. For example, during the period from 1994 to 1994 both the levels of closure and democracy were relatively low whereas between 2005 and 2013 both closure and democracy were high. Finally, the drop in the level of party system closure corresponds to the drop of the level of democracy by V-Dem after the Revolution of Dignity. The topic of the dynamics of democracy in Ukraine after the Revolution of Dignity is a contested one. For instance, contrary to the conclusions of V-Dem other indices of democracy such as The Economist Democracy Index and Freedom House state that the level of democracy has increased since 2014. In general, taking such a long-term perspective allows to evidence the explanatory power of the index of party system closure.
Graph 4. Long-term dynamics of party system closure // Oleksandr Kuchynskyi ©️
CONCLUSION

In general, regarding both indicators of party system institutionalization, they are consistent with the changes of the level of democracy across most of the period of Ukrainian independence. Except for the period from 2002 to 2006 which was highly influenced by the Orange Revolution which led to a significant increase of electoral volatility the decrease in the level of volatility was consistent with the increase in the level of democracy and vice versa. As for the party system closure, the results depend on the time perspective taken by a scholar. If the association is interpreted year-by-year, than there appears to be no association at all. Even if a 4-year perspective is taken, there again seem to be no association. However, there is no contradiction with the conceptualization of party system institutionalization as such which is a process of the stabilization of the interactions between the parties happening over a long period of time. If a longer time perspective is taken, than there is a clear association: the increase in closure is associated with democratization whereas the decrease in closure corresponds to the decrease in the level of democracy. All in all, the hypothesis about the positive association between party system institutionalization and democratization was proved regarding both indicators of party system closure (volatility and party system closure). These results can be considered robust because two indicators are based on completely different types of information used for the measurement (results of parliamentary elections for volatility and government composition for party system closure) and still lead to the same conclusion about the hypothesis.
REFERENCE LIST

Casal Bértoa, F. (2016): Database on WHO GOVERNS in Europe and beyond, PSGo.

Casal Bértoa, F., 2012. Parties, regime and cleavages: explaining party system institutionalization in East Central Europe. East European Politics , 28 (4), pp.452-472.

Casal Bértoa, F. and Enyedi, Z., 2016. Party system closure and openness: Conceptualization, operationalization and validation. Party politics , 22(3), pp.265-277.

Dix, R.H., 1992. Democratization and the institutionalization of Latin American political parties. Comparative Political Studies, 24(4), pp.488-511.

Dw.com. (2014). Verkhovna Rada approved the composition of the Cabinet of Ministers. [online] Available at: <link> [Accessed 8 Jan. 2019].

Enyedi, Z., 2016. Populist polarization and party system institutionalization: The role of party politics in de-democratization. Problems of Post-Communism, 63(4), pp.210-220.

Fedorenko, K., Rybiy, O. and Umland, A., 2016. The Ukrainian party system before and after the 2013–2014 Euromaidan. Europe-Asia Studies, 68(4), pp.609-630.

Mainwaring, S. and Zoco, E., 2007. Political sequences and the stabilization of interparty competition: electoral volatility in old and new democracies. Party politics, 13(2), pp.155-178.

McFaul, M., 2007. Ukraine imports democracy: External influences on the Orange Revolution. International Security, 32(2), pp.45-83.

Pedersen, M.N., 1979. The dynamics of European party systems: changing patterns of electoral volatility. European Journal of Political Research, 7(1), pp.1-26.

Pravda.com.ua. (2014). New ministers were shown to Maidan. Who are they?. [online] Available at: <link> [Accessed 8 Jan. 2019].

Razumkov Centre (2017). Transformation of the Party System: Ukrainian experience in a European Context. [online] Kyiv: Razumkov Centre. Available at: <link> [Accessed 8 Jan. 2019].

Rybiy, O., 2013. Party system institutionalization in Ukraine. Demokratizatsiya, 21(3), p.401.

Thames, F.C. and Robbins, J.W., 2007, August. Party System Institutionalization and the Level of Democracy. In Annual Meeting of the American Political Science Association (Vol. 29).

Unian.ua. (2015). Rada approved the resignation of Voshchevskyi from the post of a vice-prime minister. [online]
Available at: <link> .

University of Gothenburg, V-Dem Institute (2018). V-Dem Codebook. University of Gothenburg, V-Dem Institute.

Whogoverns.eu. (2015). Rules for the calculation of iPSI. [online] Available at: <link> [Accessed 8 Jan. 2019].

Www.depo.ua. (2016). Full Composition of Groysman's Government. [online] Available at: <link> [Accessed 8 Jan. 2019].